A model for the real-time forecasting of GDP in the euro area (EURO-STING)

The assessment in real time of economic developments in the very short term is a complex task. Primarily, because the macroeconomic variables needed to characterise the situation are released with a substantial lag and, occasionally, the data available are incomplete or insufficient. But also, moreover, because despite the greater abundance and more frequent availability of indicators of economic sentiment or of activity providing preliminary information on the behaviour of relevant macroeconomic aggregates, it is not immediately clear how to integrate in a structured form the occasionally contradictory information provided by these indicators. To facilitate this integration, significant progress is being made in the development of real-time forecasting models that allow the different pieces of information of relevance for forecasting economic variables to be incorporated, as and when this information is available. This article presents such a model, designed for the real-time forecasting of the GDP growth rate in the euro area: the EURO-STING (Euro Area Short Term INdicator of Growth). A description of this model is first given. Next, the main results of its estimation are described, and its forecasting power set against some of the most commonly used alternatives on the European economic forecasting scene is assessed. Finally, some brief conclusions are drawn.